CHAPTER
SIXTY-SIX
CREDIT DERIVATIVES

DOMINIC O’KANE, D.PHIL. (OXON)

Affiliated Professor of Finance
EDHEC Business School

Since the turn of the century, the credit derivatives market has become the main destination for those wishing to hedge credit risk and those looking for credit-based investments not available in the traditional cash credit markets. Two products dominate. The first is the single-name credit default swap (CDS), which transfers the default risk of a specific corporate or sovereign. The second is the CDS index, essentially a portfolio of CDS packaged into a single transaction. Together they constitute over 90% of all credit derivative transactions.

The purpose of this chapter is to set out a comprehensive description of the CDS contract, and specifically the detailed mechanics of how it works. We follow this with a detailed description of the mechanics of the CDS indices. We defer a detailed discussion of the pricing and risk-management of CDS and CDS indices to Chapter 67. However, we begin with a history of the evolution of the credit derivatives market.

EVOLUTION OF THE CREDIT DERIVATIVES MARKET

The credit derivatives market began in the early 1990s as a branch of structured finance that involved creating highly bespoke transactions customized to a particular requirement. These included hedging a credit risk, creating a structured investment product, or managing the regulatory capital of both sovereign and corporate credit risk.

By the mid 1990s the concept of a CDS contract had been established. However, the market lacked standardized legal definitions for the credit events that trigger the contract, and for the protected debt obligations. This meant that pretrade negotiations between parties often required legal assistance in order to ensure that the triggering events and the hedged debt obligations were clearly defined and understood. In addition to increasing transaction costs, this meant that trades could take several days to approve and settle.

The significant growth phase of the CDS market therefore only started following the publication of the first Credit Derivative Definitions by the International Swaps and Dealers Association (ISDA) in 1999. This document, which became known as ISDA ’99, was a response to the legal disputes that arose following the Asian crisis of 1997 and the Russian default of 1998. In addition to the new definitions, ISDA ’99 also created a short-form confirmation document that required parties to simply tick a box in order to specify the characteristics of their trade. This made it possible for the market to develop a standard contract, i.e., a contract that has a certain number of characteristics that a majority of buyers and sellers can agree on and which therefore attracts the greatest liquidity. Following its adoption of ISDA ’99, the CDS market began a period of very rapid expansion, growing from a total outstanding notional of $632 billion in 2001 to $62 trillion by the end of 2007.1

The author acknowledges the helpful comments of Samuel Morgan and Mark Ames.

A number of important changes were made to the credit derivatives market between 2000 and 2003. Many of these changes were in response to some of the credit events that had occurred2 after the end of the dot.com boom and the events of September 11, 2001, and addressed such issues as reference entity specification, guarantees, debt restructuring, deliverability, and successors. Of these, perhaps the most problematic, since it led to a regional split in the market standard, was in relation to the treatment of restructuring as a credit event. This is discussed in detail later.

The period from 2004 to 2007 saw the arrival of the CDS indices. These were adopted very rapidly by the market and became highly traded and highly liquid. They were particularly favored by credit investors who viewed them as a simple and cheap way to take a macro-level exposure to the credit markets, in much the same way as equity index futures permit investors to take a macro-level exposure to the equity markets.

In late 2003, the credit derivatives market established an automatic trade confirmation system at the Depository Trust and Clearing Corporation (DTCC) in New York, the aim being to speed up the confirmation of trades and to remove the backlog of unsigned confirmations that had built up. Then in late 2007, the DTCC also established itself as the keeper of the comprehensive trade database—the official legal record for all confirmed trades. It also set an infrastructure to automate and standardize post-trade processing.

The period 2008–2010 saw a considerable upheaval in the CDS market as regulators responded to the aftermath of the Lehman bankruptcy and the financial crisis. Despite the fact that there were $72 billion of CDS linked to Lehman Brothers registered at the DTCC at the time of its bankruptcy, settlement of the credit event went smoothly as it used a market-wide auction procedure.3 Nevertheless, this event caused regulators to become very concerned about systemic risk in the financial sector caused by the large number of outstanding contracts and a lack of transparency with respect to where the actual credit risk was sitting.4

In an attempt to reduce counterparty risk, regulators decided to require standard credit derivatives trades to be assigned to a centralized counterparty (CCP). To prepare the ground for these changes, the market introduced some new legal protocols. The first protocol, known as “Big Bang,” was introduced in April 2009 and made the following changes:

1. Creation of a market-wide determinations committee (DC) charged with determining whether or not a credit event has occurred.

2. An automatic market-wide auction mechanism for the settlement of triggered CDS contracts following all credit events with the exception of restructuring.

3. Moving the date on which the CDS credit event protection5 begins to the date 60 days before the date on which the request to consider whether a credit event has occurred is made to the determinations committee. This change was made in order to ensure that all existing contracts are triggered by a credit event even if the contract was initiated after the credit event occurred, but before the credit event had been determined.

At the same time, there was also a recouponing process in which the market switched to trading all CDS contracts on each credit with the same coupon irrespective of when the trade was done. This differed from the previous convention in which the coupon on a CDS was set at trade initiation in order to ensure that the contract had zero initial value. The market has also changed so that standard contracts now trade with a full first coupon, i.e., the first coupon starts accruing on the previous coupon payment date even if it is traded mid-period. Both changes have been made in order to increase the fungibility of contracts and to facilitate contract netting as a precursor to the centralized clearing of CDS contracts.

In late 2009, a second protocol known as “Small Bang” was adopted. This extended the “Big Bang” protocol to deal with the restructuring event as a trigger for the CDS. This protocol was intended for the European CDS market since restructuring is not a standard credit event in the North American CDS market

At the start of 2010, the DTCC formally linked itself to the Federal Reserve System where it is expected to fall under a global oversight framework involving U.S. and international regulators. The initiative to move credit derivative trades onto CCPs was well underway, especially for CDS indices, with over fifteen trillion dollars of index trades already cleared by ICE in North America and Europe by early 2011. The first CDS to be cleared via a CCP and in compliance with the new Dodd-Frank regulations was traded in February 2011. Although most CCPs are based at exchanges, trading in the CDS market trades continue to be executed in the OTC market. The CCP then steps into the trade, placing itself between the two initial parties.

As of 2011, the size of the CDS market and CDS index markets is roughly half what it was at the market peak in 2007. A significant part of this decline can be attributed to the numerous compression cycles (discussed below) that have been carried out in the last two years. Some of it may also be due to the decrease in CDS-based and CDS index–based hedging activities of more exotic credit derivatives, in particular synthetic CDOs. Indeed a snapshot of the credit derivatives market in early 2011 shows that CDS made up 57% of the gross notional of credit derivatives, whereas CDS indices made up 34%, leaving just 9% to other transaction types.6

Market Participants

Since its inception, the credit derivatives market has been dominated by commercial and investment banks. Commercial banks have used CDS to hedge the credit risk of their loan book. The advantage of CDS is that they permit the bank to hedge the risk of a specific loan privately and bilaterally. This is often preferred to selling the loan because the sale of a loan by a bank usually requires the bank to notify the borrower and such a sale may be considered to be a negative signal by the borrower and as such may be damaging to the bank–borrower relationship. Surveys of the credit derivatives market have generally found that commercial banks buy more protection than they sell as their loan book hedging requires them to be net protection buyers. Their reason for selling protection is to earn income and diversify their portfolio credit exposure.

Investment banks and the dealer arms of commercial banks play an important role in the credit derivatives markets where they act as dealers. Overall they tend to run balanced books with equal amounts of protection bought and sold.

Insurance companies are also an important user of credit derivatives, primarily as a form of investment on the asset side of their business where they tend to be net sellers of protection. Hedge funds are an important market participant as they are able to play the credit markets from a long or short position and they also tend to buy roughly as much protection as they sell. They are particularly attracted by the ability to go short and the ability to leverage as the upfront payment needed to enter a CDS contract is typically much smaller than the contract notional.

Other users of the market with a lower market share include pension funds, mutual funds and companies. Funds may be more interested in CDS indices in particular, as they provide a quick way to establish a diversified macro credit exposure onto which they can then add specific credit exposures.

The use of credit derivatives by companies is generally quite bespoke as it may be connected with a need to hedge or assume a very specific credit exposure that they may not wish to sell or cannot find in the bond market. For example, a company may use a CDS to hedge the credit risk of future receivables from a specific company.

Uses of Credit Derivatives

The considerable growth in the size of the CDS and CDS index market has been driven by its various uses. These are as follows:

Transferring credit risk: CDS are first and foremost a tool used to transfer credit risk from one party to another. For example, banks can use CDS to transfer loan credit risk off their balance sheets into the capital markets where this risk may be assumed by an investor. This creates new loan capacity on their balance sheet.

Hedge credit risk: Before the advent of CDS, it was difficult to hedge an existing credit risk by shorting corporate bonds. Buying protection using a CDS is a much easier way to achieve the same objective. CDS indices can also be used as a hedge against changes in the market-wide pricing of credit risk and also the default of individual reference entities.

Customization: As the CDS market is a bilateral OTC market, parties can structure the features of a CDS in almost any way they wish. Features that can be varied include the currency, maturity, and seniority. The more the customization deviates from the market standard, the greater the associated cost and the lower the subsequent liquidity.

Ease: A CDS index makes it simpler for an investment manager to assume an exposure to a diversified portfolio of credits in one transaction. The tight bid/offer spread also makes the cost of unwinding the position cheaper than the cash equivalent.

Leverage and yield enhancement: A CDS and a CDS index typically require only a small upfront payment. This makes it easier to leverage the underlying risk and return compared to a fully funded cash bond.

Pure credit play: The CDS and CDS index are almost a pure credit play since unlike fixed-rate corporate bonds, they have very little interest-rate sensitivity. This isolation of the credit risk is attractive for credit fund managers.

Risk decomposition: CDS can be used to hedge out specific risks in securities with multiple risks. For example, they can be used to hedge out the credit risk of convertible bonds in an attempt to isolate the economics of the embedded equity option.

Speculation: CDS enable market participants to express a positive or negative view on an underlying credit while a CDS index makes it possible to go long or short macro-level credit.

Structured credit investments: CDS can be used as the building blocks for more exotic structured credit investments that are created to provide a more tailored risk-return profile to specific types of investor.

Manage regulatory capital: Banks can use CDS to hedge credits that have a high regulatory capital charge, provided such a hedge is recognized as being economically effective by regulators.

THE CREDIT DEFAULT SWAP

The simplest way to think of the CDS is as an insurance contract in which one party, the “protection buyer,” pays a premium to be protected against a credit loss on a specific face amount of bonds or loans issued by a specified corporate or sovereign reference entity. The other party, known as the “seller of protection,” receives the premium and takes on the credit risk. Since there is a buyer and a seller of the credit risk, a CDS contract does not create credit risk, it simple transfers it from the protection buyer to the protection seller as shown in Exhibit 66–1. However, unlike an insurance contract, a CDS is a traded contract that can be valued at any time. The protection seller is any investor who takes a view that the premium is attractive given his or her view of the risk of default of the reference entity.

EXHIBIT 66–1
A CDS is a Bilateral Contract between a Protection Buyer and a Protection Seller

images

Care needs to be taken when trading CDS to be clear about to which side of the contract we are referring. In the market, “buying” a CDS contract is usually associated with buying protection. However, buying protection is economically equivalent to shorting credit risk or selling a corporate bond issued by the reference entity. Equally, “selling” a CDS contract is economically equivalent to assuming credit risk and so is akin to buying a corporate bond issued by the reference entity. Indeed we will demonstrate this equivalence in Chapter 67. To avoid misunderstandings, the term buying a CDS should always be qualified with the word protection; that is, buying CDS protection. Equally selling a CDS should be qualified as selling CDS protection. Note that this differs from the CDS index market, in which the “buyer” is usually an investor who is selling protection on the underlying portfolio of CDS. This is another reason why clarity is essential in order to prevent confusion.

CDS MECHANICS

As just stated, the purpose of the contract is to transfer the credit event risk of a specified face value amount of debt obligations issued by a stated reference entity (a corporate or sovereign) from the protection buyer to the protection seller.

The debt obligations covered by the contract are known as the deliverable obligations and these are usually bonds and loans of a specified seniority. This seniority is defined by specifying a reference obligation of the issuer with respect to which the deliverable obligations should be senior or pari passu. If a credit event occurs, then the protection buyer will receive a payment which is economically equal to the difference between the price of the impaired deliverable obligations and par.

The reason for protecting a basket of deliverable obligations rather than a specific obligation is that it enhances the liquidity of the market because the same contract can be used by hedgers to protect themselves against any one of many obligations of the issuer. If the market were to trade CDS on each specific debt obligation of the issuer, then liquidity and fungibility would be reduced and the cost of protection would almost certainly increase. However, there is a negative side effect because the protection buyer has the right to choose which bond is protected by the contract. He is therefore long a delivery option. The effect of this and how it is treated is discussed below in the section on restructuring-type credit events.

To enter into a CDS, a party trades via a dealer. This is done via the telephone or an electronic trading system. At initiation, the party usually makes an upfront payment that (depending on various factors discussed below) can be positive or negative, i.e., the payment could be from the protection seller to the protection buyer or vice versa. The model used to determine the upfront value is set out in detail in Chapter 67.

Credit protection starts on the effective date. Prior to the 2009 “Big Bang” protocols, this was the calendar date immediately following the trade date. However, under the new protocols, protection now begins 60 days before any request to consider whether a credit event occurred is made to the determinations committee. This change was made to ensure that all existing contracts would cancel out irrespective of whether they were traded yesterday or last year. This can create a rather strange scenario in which a protection buyer can buy protection after the credit event and be paid out. However, this is not as anachronistic as it may seem since the protection would have to be purchased before the credit event was determined. Also, if the information that a likely credit event had occurred was already public at the time of the purchase then the cost of the protection would have been high and close to its post-credit event determination value.

Credit protection lasts until the scheduled termination date, more commonly known as the CDS maturity date. This will fall on one of the following dates: March 20, June 20, September 20, or December 20 of a specified year. These dates are not adjusted for holidays if they fall on a weekend. They are known as the CDS Dates. What this means is that at any moment in time, the scheduled termination date of the most liquid T-year CDS contract will be on the first CDS Date which falls after the date exactly T years in the future. So, for example, on April 3, 2011, the standard five-year CDS contract has a scheduled termination date of June 20, 2016. The most liquid contracts have terms of one, three, five, seven, and ten years, with greatest liquidity focused on the five-year contract.

The Premium Leg

To pay for the credit protection, the protection buyer pays the protection seller a combination of an upfront payment (which may be negative) plus a regular fixed coupon. These payments are known as the premium leg of the CDS. Here we will discuss the fixed coupon, deferring the discussion of the upfront payment to the end of this section.

The fixed coupon is specified at contract initiation and determines the size of the quarterly regular payments from the protection buyer to the protection seller. The size of each actual cash payment is calculated by multiplying the fixed coupon by the contract notional. This is then multiplied by the year fraction between the current and previous coupon date calculated according to an actual 360 day count convention.

Before 2009, the size of this coupon depended not just on the reference entity but also on when the CDS trade was initiated. This was because the coupon was set such that the value of the CDS contract at inception was zero, thereby requiring no upfront payment. However since 2009, contracts linked to each issuer now have a fixed coupon irrespective of when they were traded. As a result, the price of the credit risk is not reflected in the size of the coupon, but in the size of the upfront payment.

The exact details of how the premium schedule is calculated are beyond the scope of this section, although more information can be found in Chapter 67. However, it is worth noting that since 2008, the market standard is for contracts to trade with a full first coupon even if they are traded between two coupon payment dates. This enables a new contract to always have exactly the same cash flows as another existing contract with the same maturity date. This is discussed with specific examples in Chapter 67.

A key feature of the premium leg is that the payments terminate following a credit event. This is shown in Exhibit 66–2. All that is paid by the protection buyer on the premium leg is the amount of coupon that has accrued between the previous coupon payment date and the credit event. This feature of a CDS means that once a credit event has occurred, the premium leg terminates and the contract can be closed out.

EXHIBIT 66–2
The Premium and Protection Legs of a CDS

images

The Protection Leg

The other leg of the CDS is the protection leg, also known as the contingent leg. This is the payment from the protection seller to cover the loss on the face value of debt, which is only triggered if there has been a credit event. Economically, the value of the payment equals par minus the recovery price of the deliverable obligations. Determination of the size of the recovery price or value is now done using a market-wide auction process that is discussed in detail below.

The market provides two mechanisms for the protection seller to effect the payment of protection. Either the protection seller can choose to pay a cash amount equal to par minus the recovery price determined by an auction, or the protection seller can request physical settlement, which means that they will pay the contract notional in cash and receive the corresponding face value of deliverable obligations at the price determined by the auction.

There are also two mechanisms for the protection buyer to receive the payment of protection. Either the protection buyer can choose to receive a cash amount equal to par minus the recovery price determined by the auction, or the protection buyer can request physical settlement, which means they will receive the notional amount in cash and deliver the corresponding face value of deliverable obligations at the price determined by the auction.

The protection buyer who elects for physical settlement can choose which deliverable obligations to deliver. They are long a “delivery option” whose value depends on the range of prices of the deliverable obligations. The protection seller who has chosen physical settlement has to accept what is delivered. This is discussed in more detail below.

The Upfront Payment

The cost of protection is paid for through both the payment of the fixed coupon on the premium leg of the CDS and through upfront payment. This is the cost which has to be paid in order to enter into the CDS contract and may be positive or negative. Whether it is positive or negative depends primarily on whether the party is buying or selling protection and how the market perceived credit quality of the reference credit compares with the fixed coupon on the CDS. The market perceived credit quality of the reference credit is captured by the CDS “quoted” spread, which is defined in detail in Chapter 67.

For example, consider a short protection CDS position on a credit that trades with a fixed coupon of 100 bp, but whose credit quality has recently deteriorated so that its CDS market quoted spread is currently 160 bp. This will require an upfront payment as the protection buyer will pay cash to the protection seller in order to offset the fact that the coupon is not high enough to compensate the protection seller for the market perceived credit risk of the reference credit. Before 2009, this contract would have had no upfront cost, as it would have paid a coupon of 160 bp.

Once a CDS has been traded, its mark-to-market value begins to change due to changes in the market view of the credit risk of the reference entity as expressed by the quoted spread and changes in interest rates.

For those familiar with the valuation of CDS contracts before the introduction of fixed coupons, the valuation and risk-management has not changed in any significant way. Once the contract has traded and the upfront payment made, the value of the contract looking forward is identical to before. For example, consider a long protection CDS contract with a coupon of 100 bp in which the current quoted spread is 160 bp. This will have a positive value for the protection buyer as he is paying only 100 bp for protection, which the market now considers to be worth 160 bp. However we do not have enough information to determine whether the position has a positive P&L (ignoring for a moment the impact of coupons paid) since this depends on the upfront value when the contract was entered into. Before the arrival of fixed coupons, the coupon on the CDS would have told us the spread when the contract was initiated and as the initial cost would have been zero, the mark to market of the CDS would, excluding coupons, be the P&L of the position.

The valuation and risk-management of a CDS is covered in Chapter 67.

CREDIT EVENTS

The credit event is the term used by the credit derivative market to define each of the events that can trigger the payment of the CDS protection. Companies can default or enter into a restructuring. The situation is even more complicated for sovereign debt since countries do not formally file for bankruptcy. Instead they can default in a number of different ways, for example, by repudiating their debt or by re-denominating their foreign currency debt. The standard credit events used in CDS need to be broad enough to encompass this range of scenarios. A list of the main such credit events is provided in Exhibit 66–3.

In Exhibit 66–3 we have stipulated which credit events are “hard” and which are “soft.” This designation refers to what happens to the debt obligations of the reference entity following the credit event. For most credit events, the debt becomes immediately due and payable. This means that all debt obligations, irrespective of maturity and coupon, which are pari passu with each other have an equal claim on the remaining value of the company and so should trade at the same price. These are what we call hard credit events.

EXHIBIT 66–3
The Main Credit Events Used in the CDS Market

images

The Restructuring Credit Event

There is only one soft credit event. This is the event of restructuring in which a company does not file for bankruptcy but agrees with its creditors a consensual restructuring of its debt obligations, but in a way that has a negative impact for the creditors.

The restructuring credit event refers to out-of-court pre-insolvency restructuring that usually requires the unanimous approval of all impaired creditors. Since the company continues to trade as a going concern, debt does not become immediately due and payable. Therefore, a restructuring credit event does not cause all pari passu debt to trade at the same price and so the debt obligations continue to trade with prices that take into account a term structure, and coupon effects (i.e., higher coupon bonds trade at a higher price than lower coupon bonds of the same maturity.

The reason why we single out soft credit events is that they present a delivery option for the protection buyer. Recall that the protection buyer can choose which obligations they wish to deliver from a basket of possible deliverables. As the deliverable obligations continue to trade at different prices caused by different coupons and maturities, there may be some value in the delivery option, and the protection buyer will always be motivated to deliver the cheapest bond or loan he can find.

As a result, if the debt obligations held by the protection buyer are not the cheapest to deliver, it will be economically preferable to sell these in order to buy the cheapest to deliver, gaining the price difference as a windfall gain. The act of always delivering the cheapest deliverable causes the protection seller to take a larger loss than if another deliverable had been chosen.

Such a situation occurred following the restructuring of the debt of the U.S. insurer Conseco Inc. in 2000. In this case, banks which held short-maturity loans that were not particularly impaired by the restructuring, and who had bought CDS protection on these loans, were able to sell these loans for close to par and use the proceeds to purchase deep discount bonds trading in the $65 to $80 range, which they could then deliver in return for par. They were therefore able to make a windfall gain at the expense of the protection sellers.

Following this, complaints were made by Conseco protection sellers. This resulted in the introduction of a new clause that formed part of the revised CDS documentation released in 2001. Called the modified-restructuring (Mod-Re) clause, it was intended to reduce the value of the delivery option following a restructuring by reducing the range of deliverable obligations in the basket. This was done by restricting the maturity of the deliverable obligations. This clause only applied to the North American CDS market.

In 2003, Europe followed by introducing its own restructuring clause. As it was a variation of the North American clause, it was known as modified-modified restructuring (Mod-Mod Re). These regional differences were due to the differing requirements of the local financial regulators and the differing importance of banks versus investors in the different regions.

We now have three different restructuring clauses, each with its own restrictions on the maturity of the deliverables. These are as follows:

Old-Restructuring (Old-Re): This is the pre-2003 standard which only imposes a 30-year maturity limit on deliverables.

Modified-Restructuring (Mod-Re): The exact maturity limits of the Mod-Re deliverable obligations are quite complicated. A rough and ready approximation is that the deliverable obligations cannot have a maturity longer than the greater of the scheduled termination date of the CDS and the restructuring date plus 30 months. If we use RMLD to denote the restructuring maturity limitation date then

RMLD = Max [TD, RD + 30 months]

where TD is the scheduled termination date of the CDS contract and RD is the date of the restructuring event.

Modified-Modified Restructuring (Mod-Mod Re): This is similar to modified restructuring. The difference is that the formula for the restructuring maturity limitation date is given by

RMLD = Max [TD, RD + A]

where A equals 60 months for restructured obligations and 30 months for nonrestructured obligations.

There is a fourth alternative which is to remove restructuring from the list of credit events. This is simply called No-Restructuring (No-Re). The situation as of early 2011 is summarized in Exhibit 66–4.

EXHIBIT 66–4
The Standard Credit Events by Region for CDS and CDS Indices

images

Settlement of a Credit Event

Ultimately, the post-credit event settlement of a CDS contract is all about determining the final recovery price for the deliverable obligations. This is done via an auction. The main objectives of the auction process are as follows:

1. To be able to handle cases in which the notional of protection bought approaches or exceeds the total outstanding notional of deliverable obligations. In this case there can be a short squeeze as protection buyers who do not own deliverables attempt to buy these obligations in order to settle their contracts. In the past this has led to the prices of deliverable obligations to rise after the credit event as protection buyers rush to buy them in order to deliver them.

2. To ensure that the recovery price is the same across the entire CDS market. This is necessary if market participants who are both long and short protection to the same maturity on a reference credit are to be sure that they will be hedged. This requires a common recovery price to be set via a public auction procedure described below.

3. To handle the soft restructuring credit event. Recall that hard credit events cause all of the pari passu debt obligations to trade at the same price. However following restructuring credit events the debt obligations continue to trade with a term structure and the value of this delivery option led to the introduction of the Mod-Re and Mod-Mod Re restructuring clauses as described earlier. The auction needs to be able to handle these restructuring clauses.

As we shall see, these requirements have all been met by the procedure that we now describe in detail.

THE CDS SETTLEMENT TIMELINE

We now set out the mechanics of settlement of a CDS contract once a credit event has occurred. Many important changes have been made to these mechanics over the past few years. To provide an overview, Exhibit 66–5 shows a timeline of the different stages that begin with credit event determination.

EXHIBIT 66–5
Timeline for a Credit Event

images

Credit Event Determination

Until 2009, the determination of whether a credit event had occurred was a bilateral decision that involved the protection buyer sending the protection seller evidence that a credit event had occurred. This evidence would consist of publicly available information such as newspaper or wire service reports of a bankruptcy, failure to pay, or restructuring.7 In most cases in which the source of the information was reputable and the events described clearly satisfied one of the specified credit events, the contract would trigger.

However, in cases in which the information was not public, was unclear, or was contested, there was a risk that different pairs of parties could come to different conclusions. This would be a problem for hedged intermediaries who would find that one contract might be triggered but the offsetting contract would not, or that the recovery rate on one contract was not the same as the recovery rate on the offsetting contract, generating an unexpected gain or loss.

To address this concern, the market introduced a new credit event determination procedure as part of the legal protocols ushered in by the 2009 Big Bang. This introduced a new body known as the determinations committee (DC) that is tasked with the job of determining whether a credit event has occurred. In actual practice there are five such DCs, one for each of the following: the Americas, Asia (excluding Japan), Japan, Australia-New Zealand, and EMEA (Europe, the Middle East, and Africa).

In order to remove any bias and to ensure that decisions are not finely balanced, the membership of each committee consists of both buy-side and sell-side representatives, and the approval of decisions of the committee requires an 80% supermajority.

The credit event determination process begins when any market participant who has signed up to the new credit derivative protocols introduced in 2009 makes a request to the secretary of the DC to investigate whether a specific reference credit has experienced a credit event. This request should be accompanied by a notice of publicly available information reporting that a credit event may have occurred. The DC then has two business days to meet, investigate, and then vote on whether a credit event has occurred. If more time is needed, they can vote to extend this time period.

Once the credit event has been determined and it is a hard credit event, all CDS contracts linked to this reference credit are automatically triggered. This is a change from the pre-Big Bang procedure in which one of the parties to the CDS would choose to trigger the contract.

However if the credit event is determined to be a restructuring, either the protection buyer or the protection seller must choose to trigger the contract. Furthermore, the exact constituents of the basket of deliverable obligations depend on which party triggered first. This is discussed in more detail below.

The rationale for not making the triggering automatic is that restructuring is not a bankruptcy event but could be a precursor to one. As a result, a protection buyer may prefer not to trigger, thereby keeping the CDS protection “alive” for a later and possibly lower recovery hard credit event.

The Compression Cycle

One of the recent innovations of the credit derivatives market has been the use of what has become known as the “compression cycle.” This is a procedure designed to enable groups of counterparties to cancel out mutually offsetting positions. This process is run periodically (typically once a quarter) across the entire credit derivatives market.

Although previously a rather ad-hoc procedure, it has also become an important stage in the settlement of a credit event, where it is run across the relevant CDS contracts after it has been determined that a credit event has occurred and before the auction begins. However, this is only done for those contracts held by dealers who choose to participate in the compression cycle. The effect of running the compression cycle is to reduce the number of contracts that have to participate in the subsequent auction. This will reduce the amount of physical assets needed to settle the credit event.

The main provider of this compression service is a company called TriOptima. A third party is used to provide this service because compression is all about the cancellation of sets of contracts between three or more dealers. To do this, the parties either need to reveal their full positions list to each other or they need to involve a neutral third party. Dealers prefer the latter, and so introduce a third party to collect the information on the condition that they do not disclose position data to the other parties.

An algorithm is run that calculates the specific set of transactions that if executed would most reduce the outstanding number of contracts. These transactions are then reported back to the dealers who then can choose whether or not to execute them. Between 2008 and 2010, this compression process eliminated around $14.5 trillion of credit derivative contracts from the market.

Deliverables for a Nonrestructuring Credit Event

Within a short period of the credit event being determined by the DC to be a hard credit event, the DC publishes a timetable of when the credit event was determined, when the auction will be held and when the auction will be settled. A list of the deliverable obligations is also published.8 For hard credit events, there is just one auction as there is just one basket of deliverables for all CDS contracts since the maturity limitation is at 30 years.

Deliverables for a Restructuring Credit Event

When the credit event is a restructuring, the situation depends on the region. In North America, there is no restructuring credit event in standard contracts and so there is only one auction. In Asia, the use of the old-restructuring convention means that there is only one basket for all contracts. In Europe, the situation is more complicated. The market standard is to include a restructuring credit event. However the standard restructuring clause is modified-modified restructuring in which the basket of deliverables depends on the scheduled termination date of the CDS contract.

The purpose of the standard modified-modified restructuring clause found in European CDS contracts is to limit the basket of deliverables in order to reduce the value of the delivery option. In approximate terms, the restructuring maturity limitation date for a deliverable into a CDS contract with a scheduled termination date (TD) is given by

RMLD = Max [TD, RD + 60 Months]

for restructured obligations, and

RMLD = Max [TD, RD + 30 Months]

for nonrestructured obligations. In both cases the basket of deliverables depends mainly on the TD of each CDS contract.

Reconciling this restructuring clause with an auction mechanism presents an immediate difficulty. If the reference credit has a large number of deliverable obligations distributed across the maturity spectrum, it is possible that there will be a different basket of deliverable obligations for each possible CDS scheduled termination date. Given that the longest maturity CDS contracts have a scheduled termination date in around ten years, and given that CDS scheduled termination dates are quarterly, this implies that we could have more than 30 distinct baskets of deliverables9 for a given reference entity. However, holding this many auctions in parallel would be organizationally burdensome. It would also split the universe of relevant contracts across many auctions, resulting in a loss of liquidity and price competition in each auction.

To overcome this problem, the market has decided to run a maximum of eight different auctions by assigning each CDS contract to one of eight maturity buckets. The buckets and what they contain are listed in Exhibit 66–6. In practice, fewer than eight buckets may be used if a lack of deliverables across the maturity spectrum means that two or more buckets are identical. In this case only the earliest of these buckets will have an auction. In addition, a minimum liquidity requirement states that an auction will only be held if 300 or more contracts linked to that bucket are triggered and if there are five or more dealers who are party to these transactions.

EXHIBIT 66–6
The Maximum Set of Auction Buckets Used after a Restructuring Credit Event


Bucket


Nonrestructured debt obligations out to 30 months and restructured debt obligations out to 5 years;

All deliverable obligations to 5 years;

All deliverable obligations to 7.5 years;

All deliverable obligations to 10 years;

All deliverable obligations to 12.5 years;

All deliverable obligations to 15 years;

All deliverable obligations to 20 years; and

All deliverable obligations to 30 years.


If a bucket is not used, then the protection buyer and seller have a movement option. This allows the protection buyer to choose to move the contract to use the next earliest maturity bucket. The protection seller also has a movement option and can also choose to move the contract to the longest maturity 30-year bucket. If both parties wish to exercise the movement option, then the protection buyer’s decision is the one that takes priority.

Prior to the auction, each CDS contract must be assigned to one of these buckets. To simplify matters, one of the duties performed by the DC is to calculate and publish a table that maps ranges of scheduled termination dates to specific buckets. However, there is one final step before CDS holders know which maturity bucket, and hence which auction, is relevant to their contract. This is the triggering decision.

Restructuring: Triggering a Credit Default Swap

Unlike hard credit events, the triggering of a contract following a restructuring credit event is not automatic. In fact it is actually asymmetric in the sense that the basket of deliverable obligations depends on which party, if any, is the first to trigger the contract. The rules are as follows:

• If the contract is triggered by the protection buyer, then the basket that applies is the one that corresponds to the scheduled termination date of the CDS contract.

• If the contract is triggered by the protection seller, then the basket that applies is the longest maturity (30-year) basket.

The consequence of these rules is that each party wants the other to trigger. To see why, consider the holder of five-year CDS protection. He will prefer the protection seller to trigger because the 30-year bucket contains more deliverable obligations than the five-year bucket, and so the value of the delivery option is greater. At the same time the protection seller wants the protection buyer to trigger since he does not wish to be delivered a long-dated obligation.

However, the protection buyer also knows that if he fails to trigger within the specified period, the ability to trigger for this credit event is lost. This is known as use it or lose it. Suffice to say that both protection buyers and sellers will usually wait until the last moment to trigger. However, great care needs to be taken not to miss the triggering deadline altogether.

There is another complicating factor, which is that the decision period for the protection buyer and seller are not always the same. Typically these periods begin together, but the protection buyer has five days, whereas the protection seller has three days to decide.10

If the contract is not triggered by either party, it does not follow that the protection is lost altogether. It just means that it cannot be used for this credit event but can be used for any subsequent credit event.

The Auction Process

Before the auction, holders of CDS contracts need to decide whether they wish to use cash or physical settlement. A protection buyer who elects for physical settlement will end up with par in cash and a short position in the deliverable obligation after the auction. This replicates the economics of delivering a bond to the protection seller in return for par. Equally a protection seller who elects for physical settlement will end up paying par in cash and owning a deliverable obligation. This replicates the economics of paying par and taking delivery of a deliverable obligation from the protection buyer.

The auction breaks the link between the protection buyer and seller since each party to a CDS can each choose a settlement mechanism that may be different to that of their counterparty.

Prior to the auction, market participants who wish to have physical settlement are able to submit physical settlement requests via their dealers. These must arrive before 5 p.m. on the day before the auction. Each request must be in line with the market participant’s net CDS position (i.e., a net long protection position must correspond with a request to sell bonds, and vice versa).

A protection seller will choose physical rather than cash settlement in order to realize a view that the current recovery price is too low and will rise after the auction. This view requires the protection seller to hold the physical asset.

The auction is a two-stage process that is administered by Creditex and Markit. It is carried out via the Creditex inter-dealer electronic trading platform. This means that non-dealers cannot participate directly but must do so via their chosen dealer. Both stages occur on the same auction date, with stage one usually occurring in the morning and stage two in the afternoon.

Auction: Stage One

Stage one of the auction is used to determine an indicative recovery price for the deliverable obligations and the net open interest to buy or sell physical assets that will be used in stage two of the auction. Stage one is limited to parties who have CDS positions and have submitted physical settlement requests.

The process is short, lasting only 15 minutes. In this time, the participating dealers submit two-way bid/offer prices for the deliverable obligations. At the same time, the dealers submit physical settlement requests on behalf of their customers and on their own behalf. Note that there is a maximum limit on the bid/offer spread submitted by dealers that prevents dealers from making uncompetitive markets. There is also a minimum trade size. Both of these constraints are set in advance by the DC and are a function of the liquidity and market outstanding notional of the reference credit.

At the end of this 15-minute period, the auction administrator has a further 15-minute period during which they calculate and then publish the indicative recovery price. This is more formally known as the internal market midpoint price (IMMP) and is calculated using an algorithm that rejects outlying price combinations (to prevent manipulation) and averages the remaining bid/offer spreads that are then rounded to the nearest 1/8th of a point. The internal market midpoint price will then be used in stage two of the auction procedure to constrain the final auction price.

The final calculation is the net open interest, which is simply the aggregation of the physical settlement requests. This establishes the size and direction of the demand of market participants to use physical settlement, i.e., the total amount of deliverable obligations to be traded in stage two of the auction and whether the net open interest is to buy or sell these deliverable obligations.11

Auction: Stage Two

Stage two of the auction occurs when the parties bid to satisfy the net open interest determined in stage one. This process results in a market-wide recovery price that is then used to settle all CDS contracts, including both those that are cash and physically settled.

Stage two begins two or three hours after the end of stage one, thereby giving market participants time to decide how to respond to the results of stage one. Unlike stage one it is an auction that is open to all parties who have signed up to the ISDA protocol, not just those who hold a CDS position in the relevant reference entity. This maximizes the number of participants and should maximize the chances of a competitive auction.

If the net open interest of stage one of the auction is to sell bonds, then the aim of stage two is to match this open interest with those who wish to buy the bonds. If the net open interest of stage two of the auction is to buy bonds, then the aim of stage two of the auction is to find bond sellers.

At the start of stage two, auction participants have a 15-minute window during which they submit limit orders of a size that is only limited by the size of the net open interest. The order prices can be at whatever price they decide. To do so, they have to commit that they will be willing to transact the limit order size of deliverable obligations at the final auction price.

When this period is over, an order matching algorithm is run on all of the submitted orders. If the open interest was to buy bonds, then the auction is one in which participants submit offer prices to sell bonds. The open interest is then filled starting at the lowest price until all of the open interest has been filled. The price at which this happens is the final auction price. If the open interest was to sell bonds, then the auction is one in which participants submit bid prices to buy bonds. The open interest is then filled starting at the highest price until all of the open interest has been filled. The price at which this happens is the final auction price.

This final auction price may not become the official final auction price, as there is a condition that requires that the final price be within a certain range, known as the cap amount, of the IMMP calculated in stage one. If the final auction price is above the IMMP by more than the cap amount, then it is set equal to the IMMP plus the cap amount. If the final auction price is below the IMMP by more than the cap amount, then it is set equal to the IMMP minus the cap price.

After the auction, dealers are paired off with each other so that the demand to buy or sell deliverable obligations that were submitted in the first stage of the auction can be satisfied by the demand to satisfy the open interest that was filled by the orders submitted in the second stage of the auction. The dealers then in turn enter into trades with their customers to satisfy their physical settlement requests, and the settlement procedure is then complete.

CDS INDICES

Following the CDS, the most important product in the credit derivatives market is the CDS index. Unlike CDS, CDS indices are primarily used by investors to easily obtain a diversified credit exposure as they enable an investor to take a broad exposure to the credit market in a single transaction. The upfront cost required to enter into a CDS index is usually a small percentage of the notional and this makes it possible to leverage the risk of a CDS position. CDS indices are also used as a way to hedge the credit risk associated with a broad portfolio of credit names. This can be useful for credit funds, hedge funds, or banks wishing to hedge the so-called systemic market credit risk.

The first CDS indices appeared in 2002 to the extent that there were competing index providers in each regional market. This situation persisted until the middle of 2004. At this time the North American indices fell under the name CDX, while their European and Asian equivalents fell under the name iTraxx. The main CDS indices that now trade are listed in Exhibit 66–7. Of these the most liquid are the North American and European investment-grade indices, known as CDX IG and iTraxx Europe, respectively.

Note that in the CDS index market, “buying the index” means assuming the credit risk by selling protection and receiving the coupon. This differs from the CDS market, in which “buying” generally means buying protection. It is worth noting that the CDS indices are equally weighted when issued. This differs from stock indices, which are either market cap-weighted or price-weighted. It also differs from the traditional fixed income benchmark indices found in the credit markets, which are almost always weighted by the outstanding debt of each issuer. This was done for reasons of simplicity and to maximize diversification. However, this can change during the life of an index if one or more companies in the index splits (de-merges) or two or more of the companies in the index merge.

CDS Index Constituents

A CDS index is essentially a portfolio of CDS in which the credits in the index satisfy a certain set of inclusion criteria. For example, the CDX IG index is the CDS index, which contains the 125 most liquid investment-grade (IG) credits that trade in the CDS market and are domiciled in North America. The main inclusion criteria for the main CDS indices are shown in Exhibit 66–7.12

EXHIBIT 66–7
The Standard CDS Indices Showing the Type and Number of Credits in Each

images

At any moment in time, there is not just one of each of the CDS indices trading in the market but several series. This is because a new series of each CDS index is issued every six months. Although the new version of the index will have the same credit inclusion criteria as the whole series, the actual credit names of each series of the index may differ if certain credits were upgraded or downgraded, defaulted, or became less liquid since the previous issue date. The most recently issued series of any index is known as the “on-the-run” and is usually the most liquid.

There is also a further subdivision of a CDS index series by maturity. On the issue date, new series of a specific index are issued with terms which will typically include one, three, five, seven, and ten years. The maturity date of the T-year index will be on one of the standard CDS Dates—March 20, June 20, September 20, and December 20.

For a series that rolls on March 20 and September 20 each year, the corresponding maturity dates will be June 20 and December 20 T years later. This means that the time to maturity of the term T index will actually be T years plus three months at issue and this will have fallen to T years minus three months by the time the next series of the index is issued six months later. This allows an investor to continuously roll his index position into the next “on-the-run” index and so be able to maintain his average maturity at close to T years.

CDS Index Mechanics

To enter into a CDS index, the buyer (or protection seller) pays an upfront cost. The value of this may be positive or negative, and so may actually result in a payment to the index buyer. This depends on the size of the fixed coupon paid by the index and the current trading level of the index.

Each series of a CDS index has its coupon set on the issue date, and this coupon is fixed thereafter. It is usually set at a value that makes the initial upfront value of the index close to zero. Once the index series has been issued, changes in the market perceived credit quality of the index causes the upfront value to change. This is exactly analogous to the mechanics of the CDS market, in which contracts linked to each specific reference entity adopted a fixed coupon in 2009.

The economics of a CDS index are identical to those of a portfolio of CDS contracts where the coupons on the CDS contracts equal the index coupon. As with a CDS, there is both a premium leg and a protection leg.

The Premium Leg

The premium leg of a CDS index is the payment of the coupon to the buyer of the index. The actual payment amounts are determined by multiplying the index fixed coupon by the day count fraction13 that has elapsed since the previous coupon date and then by the surviving notional of the index reference portfolio. The surviving notional at initiation is the contract notional. However, over time this may change if there are credit events.

As credits in the reference portfolio experience credit events, the surviving notional of the index is reduced in proportion to the number of credits remaining. So if two credits in a reference portfolio of initially 125 credits experience a credit event, the surviving notional of the index is 123/125 times the initial notional and the size of the coupon is also reduced to 123/125 times the initial coupon. Note that the notional does not depend on the size of the recovery rate of the defaulted credits, just the number of credit events.

The economics of a CDS index premium leg are therefore equivalent to that of a portfolio of CDS premium legs linked to each of the reference credits in the index, which all pay a fixed coupon equal to the index coupon. The premium leg is paid until the end of the contract, the exception being the unlikely case of all of the credits in the reference portfolio experiencing a credit event.

The Protection Leg

The protection leg of the CDS index is identical to that of an equally weighted portfolio of CDS protection legs on the reference portfolio. If a credit event occurs on a reference credit in the index, the index buyer must pay the index seller the loss on that credit. The process is as follows:

1. The credit is immediately removed from that index series across all maturities.

2. The new index without the credit is thereafter assigned a new version number to distinguish it from earlier versions of the index series.

3. The reference credit that was removed then becomes a standalone CDS position with the notional corresponding to its notional size in the index, e.g., if the index position size was $10 million and there were 125 credits in the index, then the notional of the standalone CDS contract will be $80,000.

The standalone CDS position then takes part in the standard auction procedure as described earlier. In economic terms, this will result in the index buyer having to pay the index seller par minus the auction recovery rate on the CDS notional.

There is an added complication if the credit event was a restructuring. In this case, the index buyer and seller have the option to trigger the standalone contract. If they both choose not to trigger, then the standalone CDS position continues to sit on their respective books until maturity or until it is triggered by a subsequent credit event. However, this is usually only a consideration for European and Asian index holders since, as shown in Exhibit 66–4, the standard for North American indices is to exclude restructuring as a credit event.

IMPORTANCE OF THE CDS MARKET

The credit derivatives market has become an important branch of the credit markets. The ease with which a single-name CDS can be traded has made it an invaluable hedging tool for many, most notably bank loan risk managers. The CDS index has also become an important investment tool for credit investors seeking to quickly and easily obtain a macro-level credit exposure.

In the aftermath of the financial crisis of 2008, the market quickly responded with a number of fundamental changes in the format and mechanics of the contract to increase fungibility and so decrease counterparty risk. The triggering and settlement of credit events were revised to ensure that there is a common market-wide process and to minimize the risk of failing to settle.

Many of the changes were intended to facilitate making all market participants face centralized counterparties (CCPs) rather than dealers in an attempt to reduce the systemic risk associated with the default of a major dealer. This process began in 2009 and by February 2011, one of the major clearing houses, the ICE Trust in North America and ICE Clear in Europe, had together cleared more than $15 trillion of credit derivatives, of which over 80% were CDS indices.

One area of valid criticism of the market was its opaqueness pre-2008, which was unsurprising given the OTC nature of credit derivative transactions. The publication of high-level position data by the New York-based DTCC who are responsible for CDS settlements has added a new level of transparency with respect to market size, the reference entities being traded and the different types of counterparties.

A recurring pattern of the credit derivative market has been that each time there is a credit crisis that exposes any flaws in the mechanics of the market, the market moves quickly to attempt to eliminate such flaws so that the same problems will not recur. This reflects the importance that the market attributes to a well-functioning credit derivatives market. As a result, the market is now better settled, more transparent, less exposed to systemic risk, and better regulated than it has ever been.

KEY POINTS

• The credit derivatives market is dominated by the single-name credit default swap (CDS) and the CDS indices.

• The growth of the CDS and CDS index market has been driven by its many uses. These include transferring credit risk, hedging credit risk, isolation of credit risk, leveraging and yield enhancement, structured credit investments, and managing regulatory risk.

• The simplest way to think of the CDS is as an insurance contract in which one party, the “protection buyer,” pays a premium to be protected against a credit loss on a specific face amount of bonds or loans issued by a specified corporate or sovereign reference entity.

• The other party to the CDS, known as the “seller of protection,” receives the premium and takes on the credit risk. The protection seller makes a contractual payment to the protection buyer if a credit event occurs during the tenure of the contract.

• The “credit event” is the term used by the credit derivative market to define each of the events that can trigger the payment of the CDS protection. Credit events may include bankruptcy, failure to pay, obligations acceleration, repudiation/moratorium of debt, and restructuring of debt.

• The event of restructuring is a “soft” credit event since debt can continue to trade with a term structure after a restructuring. Since this can create a potentially valuable “delivery option” for protection buyers, it is excluded from the standard CDS contract on North America-domiciled companies. However, regulatory and other differences mean that it is included in standard contracts on Europe- and Asia-domiciled companies.

• To avoid a “short squeeze” following a credit event and to ensure that the same recovery rate is used across the market, the CDS market has introduced an auction process in which holders of the relevant CDS engage and only the net amount of deliverable obligations need to be physically traded.

• Credit default swap indices are equally weighted portfolios of up to 125 CDS that can be traded in one single highly liquid transaction. The CDS index market provides asset managers with a highly liquid way to assume or hedge an exposure to market-wide or sector-wide credit risk.

• Following the financial crisis of 2008–2009, the credit derivative market made a number of substantial reforms in order to ensure a well ordered market-wide settlement mechanism following credit events, to improve transparency and reduce systemic risk via the use of centralized counterparties.

• A CDS or CDS index contract can be sold at any time before maturity, allowing the owner to realize any gain or loss resulting from changes in the market perceived credit quality of the reference entity.

..................Content has been hidden....................

You can't read the all page of ebook, please click here login for view all page.
Reset
54.236.245.71